Skip to main content

Table 1 Running 10-fold cross-validation on the iDASH dataset with 1579 samples and 18 selected genotypes

From: Logistic regression over encrypted data from fully homomorphic encryption

Training method # iterations Avg. training time Avg. AUC Avg. AUC (unencrypted)
GD + σ3 36 115.33 h 0.690 0.690
1-Bit GD + σ1 36 14.90 h 0.668 0.690
  1. The first average AUC value is obtained from running the training algorithm using SEAL on encrypted data. The second AUC value is obtained from running the same algorithm on unencrypted data using MATLAB